For some fifteen years now, I have been using Circuitmaker for simulation purposes. I like it and have learned to work really fast with it. I have tried to migrate to some more modern software, but I'm too impatient to learn how to use them.
The thing with CM is that the spice engine is not too reliable, it seems. The usual errors such as gmins may be dealt with by adding resistors here and there. But some other things are totally crappy and are obviously bugs.
Such as:
@6 : no such vector.
or
warning: fourier line too long.
or
Simply the spice engine crashes and takes down the whole application
Now my question to you folk's is the following:
Are modern simulating tools more reliable?
The thing with CM is that the spice engine is not too reliable, it seems. The usual errors such as gmins may be dealt with by adding resistors here and there. But some other things are totally crappy and are obviously bugs.
Such as:
@6 : no such vector.
or
warning: fourier line too long.
or
Simply the spice engine crashes and takes down the whole application
Now my question to you folk's is the following:
Are modern simulating tools more reliable?
I do not know this sw, I am using LTspice like most of the other DIYers. Works stable with Windows and even Linux/Wine. My experience es that in many cases the bug sits before the screen. You can draw quite simple circuits that run into problems, most of the time "ideal" circuits that are far from reality. Inserting small resistors is an appropriate fix to make them more realistic and makes simulation affordable.
OK, most diy-ers use LTspice. Do you encounter these obvious bugs that Circuitmaker has? The ones that makes you want to throw out the computer through the window?
My simulator seems to have some sinister conscious. When I'm starting the project and I'm working slowly and methodical, the software works flawless. But as soon I'm getting eager and excited the thing starts to plague me.
And every time I want to make a very crucial measurement, the one that really everything depends on, then usually the message "@6 : no such vector" comes up. From that point everything I do with the circuit just greets me with that damned message. It' evil, my simulator, that's it.
My simulator seems to have some sinister conscious. When I'm starting the project and I'm working slowly and methodical, the software works flawless. But as soon I'm getting eager and excited the thing starts to plague me.
And every time I want to make a very crucial measurement, the one that really everything depends on, then usually the message "@6 : no such vector" comes up. From that point everything I do with the circuit just greets me with that damned message. It' evil, my simulator, that's it.
Hi, I forgot to say anything about Spice engine base simulators in your Godzilla amp thread under the Solid State section.
Unfortunately my experiences are very old and goes back many years when I used Orcad PSice at work, but haven't used any since then, except the very popular LTspice, which I installed some years ago, but I ended up using it only for drawing up lot of schematics.
I just remember that some circuits won't simulate because the program didn't understand some nodes, I then just added a huge resistor from that node to ground or vcc, say 10-100 MOhm which wouldn't impact on the circuits functionality but it helped the simulator figure out how to calculate a node.
I can't remember how those fault messages looked like when I used the simulator, but out of curiosity I searched a bit on the error messages you are reporting and found a few but all related to different type of Spice engine based simulators such as NGspice, Qucs etc..
I suspect that, because those different simulators are all just an interface between the Spice engine and the user, I am thinking here that some user interfaces handles error messages differently, perhaps CM may not be up to the task handling error messages in a sane manner but instead crashes.
As soon as I have a better computer I am going to install LTspice again and learn to use it properly, it is widely used and constantly updated so it's a safe choice I think.
ps. Sorry to see you leave the Godzilla project, perhaps those dreaded BJT hFE anomalies scared you away. 😉TOP
Unfortunately my experiences are very old and goes back many years when I used Orcad PSice at work, but haven't used any since then, except the very popular LTspice, which I installed some years ago, but I ended up using it only for drawing up lot of schematics.
I just remember that some circuits won't simulate because the program didn't understand some nodes, I then just added a huge resistor from that node to ground or vcc, say 10-100 MOhm which wouldn't impact on the circuits functionality but it helped the simulator figure out how to calculate a node.
I can't remember how those fault messages looked like when I used the simulator, but out of curiosity I searched a bit on the error messages you are reporting and found a few but all related to different type of Spice engine based simulators such as NGspice, Qucs etc..
I suspect that, because those different simulators are all just an interface between the Spice engine and the user, I am thinking here that some user interfaces handles error messages differently, perhaps CM may not be up to the task handling error messages in a sane manner but instead crashes.
As soon as I have a better computer I am going to install LTspice again and learn to use it properly, it is widely used and constantly updated so it's a safe choice I think.
ps. Sorry to see you leave the Godzilla project, perhaps those dreaded BJT hFE anomalies scared you away. 😉TOP
- Status
- Not open for further replies.